A Novel Methodology For Fault Detection, Classification And Location In Transmission System Based On Dwt & Anfis

A novel and precise numerical relaying approach based on discrete wavelet transform (DWT), Artificial Neural Network (ANN) and Adaptive Neuro-Fuzzy Inference System (ANFIS) is presented in this paper. Detection and categorisation of the faults are achieved using designed neural network (NN) which is trained by samples of voltage and current, procured from the simulated test system. However, for localizing the faults, ANFIS is employed. The summation values of discrete wavelet coefficients of 3-phases are extracted for training and testing of ANFIS model. Daubechies db4 (mother wavelet) is applied for decomposing 3-phase current samples. For validating the effectiveness of the present approach, it is checked on a 400 kV transmission system in MATLAB. The results obtained demonstrate that proposed method is effective in detecting, classifying and predicting faults position in the network.